Oral squamous cell carcinoma (OSCC) is a worldwide problem. An estimated 30,000 people in the United States are diagnosed with OSCC each year. Since the majority of OSCC develop from precursor lesions, accurate identification and management at the precursor stage offers the best hope at reducing OSCC morbidity and mortality. Leukoplakias (white patches) are the most common precursor lesion of OSCC. The malignant transformation rate of oral leukoplakia is variable with rates as high as 31.4%. No reliable mechanism exists to selectively identify which leukoplakias will undergo malignant transformation. Standard clinical practice relies on the microscopic detection of high-grade epithelial dysplasia to dictate which leukoplakias are subject to treatment. The problem with this approach is that it fails to account for the subset of leukoplakias that are histologically nondysplastic and low grade yet progress to cancer. It has been reported that as high as 16% of nondyplastic leukoplakias progress to OSCC. Developing a predictive clinical modality that can accurately identify progressive lesions among clinical leukoplakias is in critical need. Once identified, these lesions can receive appropriate clinical management and follow-up, thereby halting malignant transformation and significantly improving the clinical outcome. Early detection and management of progressive leukoplakias will significantly contribute to eradication of OSCC. The aim of the proposed study is to develop and validate a microRNA (miR) marker-based predictive modality for oral cancer progression. In the initial phase of the study, we propose to perform a genome-wide miR expression assessment via next generation sequencing in 10 incisional biopsy tissue samples from patients with non dysplastic or low grade dysplastic oral leukoplakias who had 5 year disease-free survival (Group 1), compared with 10 from age- and gender- matched patients with non dysplastic or low grade oral leukoplakias that progressed to OSCC within 5 years (Group 2). Bioinformatics analysis will be performed to integrate sequencing data to identify miR target genes and miR-mediated oncogenic networks and pathways. Top miR candidates that can identify those patients with progressive lesions with high sensitivity and specificity will be selected to form a prognostic miR marker panel in Aim 1. The candidate marker panel will then be validated in Aim 2 in an additional 40 Group 1-Group 2 pairs using RT-qPCR. To our knowledge the proposed innovative approach has never been applied to predict progression of oral leukoplakias. This proposal offers scalability and cost containment while allowing an agnostic approach to discovery of molecular changes across the genome that is followed by independent validation.